feat: random_ggplot('ts'), random_mock(), random_text(var=) — issues #4 #1 #3 by VincentGuyader · Pull Request #16 · ThinkR-open/shinipsum

VincentGuyader · 2026-05-12T07:57:42Z

Summary

#	Type	Résumé
#4	feat	`random_ggplot()` gains a `"ts"` type (alias `"timeseries"`) returning a time-series oriented plot, with a `Date` on the x axis. 3 variants (`economics$unemploy` line, `AirPassengers` line+point, `economics$psavert` area). `"ts"` is part of the random pool.
#1	feat	New exported `random_mock(widths, heights = 3, labels = NULL)`: builds a quick "zoning" of a Shiny UI as a row of bordered, labelled bootstrap columns, returned as a self-contained `htmltools::tagList` (CSS `<style>` + columns) droppable inside `fluidPage()` / `bslib::page_fluid()`. Default labels are short random lorem snippets.
#3	feat	`random_text()` gains a `var` integer argument: `var = 1` keeps the historical output byte-for-byte, higher values return distinct (wrapping) slices of the lorem corpus — useful for repeated calls in a list.
#13	fix	`random_ggplot("line")` no longer triggers ggplot2's `size → linewidth` deprecation warning. (Also addressed in PR #15; trivial to reconcile.)

New dependency: htmltools in Imports (already a transitive dep via plotly / DT / dygraphs).

Test plan (TDD, red first)

New test files written before the implementation and watched fail for the right reason:
- tests/testthat/test-ggplot-ts.R — "ts"/"timeseries" returns a ggplot whose data has a Date date column, build emits no warning, "ts" in the random pool.
- tests/testthat/test-mock.R — tagList of bootstrap columns, custom labels, recycling of heights/labels, auto labels, widths validation.
- tests/testthat/test-text-var.R — var = 1 ≡ historical behaviour, distinct var ⇒ distinct text, size preserved, large var wraps (no NA), combines with offset, validation.
Added a coverage test for random_ggplotly() (previously uncovered).
devtools::test(): 3398 PASS / 0 FAIL / 0 WARN.
R CMD check --no-manual (tarball): 0 error / 0 warning / 1 note (utils import unused — pre-existing on master).
Package coverage: 98.07% → 98.76%; R/Plot.R, R/Mock.R, R/Text.R at 100%.

Tracking notes in dev/SUIVI_AUTONOME.md.

Closes #1
Closes #3
Closes #4

Addresses pr-reviewer nits on #16: explicit heights guard with tests, and distinct wrapping default labels via random_text(var = i).

VincentGuyader · 2026-05-12T08:19:13Z

CI status: R-CMD-check is green on all 5 platforms (macOS, Windows, Ubuntu devel/release/oldrel-1), 0 error / 0 warning.

The only red check is test-coverage, which fails at setup with This request has been automatically failed because it uses a deprecated version of actions/cache: v2 — that workflow (.github/workflows/test-coverage.yaml) is still on r-lib/actions/...@v1 + actions/cache@v2. This is pre-existing on master and on every branch; modernising the workflows is already handled by #14 / #15, so I deliberately did not duplicate that change here.

Ready for review.

Copilot

Pull request overview

Adds new prototyping helpers and improves existing generators in {shinipsum}: time-series ggplots, UI “zoning” mocks for Shiny layouts, and deterministic text variants for repeated placeholder text use.

Changes:

Extend random_ggplot() with a new "ts" / "timeseries" type and include it in the random pool; also remove ggplot2 size→linewidth deprecation warning for the "line" type.
Add new exported random_mock() to generate a bootstrap-row mock UI layout as a self-contained htmltools::tagList().
Extend random_text() with a var argument for repeatable “variant” slices of the lorem corpus, plus new tests and documentation updates.

Reviewed changes

Copilot reviewed 13 out of 17 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
`R/Plot.R`	Adds `"ts"` time-series ggplot variants, alias handling, and updates `geom_line()` to use `linewidth`.
`R/Text.R`	Adds `var` argument with validation and variant/wrapping logic for both char- and word-based outputs.
`R/Mock.R`	Introduces `random_mock()` and internal CSS helper for Shiny UI layout zoning.
`R/globals.R`	Extends `globalVariables()` to cover new NSE columns used by the `"ts"` plots.
`tests/testthat/test-ggplot-ts.R`	Adds tests for `"ts"`/`"timeseries"` behavior, Date x-axis, no-build-warnings, and pool inclusion.
`tests/testthat/test-ggplot.R`	Adds deterministic `random_ggplotly()` coverage test using non-optional plot types.
`tests/testthat/test-mock.R`	Adds tests for `random_mock()` output structure, recycling, auto labels, and input validation.
`tests/testthat/test-text-var.R`	Adds tests for `random_text(var=)` backward compatibility, variation, wrapping, and validation.
`DESCRIPTION`	Adds `htmltools` to Imports; adds `MASS`/`hexbin` to Suggests; updates `RoxygenNote`.
`NAMESPACE`	Exports `random_mock`; imports new symbols from `attempt`, `ggplot2`, and `htmltools`.
`NEWS.md`	Documents the new features and the ggplot2 deprecation warning fix.
`man/random_ggplot.Rd`	Updates docs to include `"ts"` and its behavior.
`man/random_text.Rd`	Updates docs to include the new `var` argument.
`man/random_mock.Rd`	Adds documentation for the new `random_mock()` function.
`man/shinipsum-package.Rd`	Minor formatting adjustment in package description.
`.Rbuildignore`	Adds `dev/` to build ignore patterns.
`dev/SUIVI_AUTONOME.md`	Adds internal development tracking notes for the implemented issues.

Files not reviewed (4)

man/random_ggplot.Rd: Language not supported
man/random_mock.Rd: Language not supported
man/random_text.Rd: Language not supported
man/shinipsum-package.Rd: Language not supported

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

@@ -21,18 +26,38 @@ random_text <- function(nchars = NULL, nwords = NULL, offset=0){
    ~ !is.null(.x),
    "You can't chose both nchars and nwords"


+    } else {
+      n <- length(shinipsum::lorem_words)
+      shift <- (var - 1L) * nwords
+      idx <- ((1 + offset:nwords + offset + shift) - 1L) %% n + 1L


@@ -18,7 +18,10 @@ random_ggplot <- function(type = c("random", "point", "bar",
                                   "density", "density_2d", "dotplot",
                                   "hex", "freqpoly", "histogram",
                                   "ribbon", "raster", "tile",


- typo 'chose' -> 'choose' in random_text() error message - rewrite random_text() var>1 word-indexing without ambiguous +/: precedence - de-duplicate 'tile' in random_ggplot() type choices (was biasing the random pool) + add a no-duplicates test

VincentGuyader · 2026-05-12T11:02:40Z

Thanks @copilot — all three points addressed in dce10a3:

`R/Text.R` — "chose" → "choose": fixed (it was a pre-existing typo in the nchars / nwords mutual-exclusion error message).
`R/Text.R` — ambiguous `+`/`:` precedence for `var > 1`: the word-based branch is rewritten explicitly — start <- (offset + (var - 1L) * nwords) %% n; idx <- ((start + seq.int(0L, nwords)) %% n) + 1L. It now returns a consistent nwords + 1 words for any var > 1, with wrap-around, and var == 1 is still byte-for-byte the historical output.
`R/Plot.R` — "tile" listed twice in `type`: de-duplicated in both the formals and the roxygen @param, which removes the sampling bias for type = "random". Added a regression test asserting anyDuplicated(eval(formals(random_ggplot)$type)) == 0. (The unreachable "tile" = sample(160:161, 1) switch branch is left untouched — pre-existing, harmless, out of scope here.)

Re: the red test-coverage check — that's the pre-existing actions/cache@v2 auto-fail in the old workflow, fixed by #14; unrelated to this PR.

Copilot

Pull request overview

Copilot reviewed 13 out of 17 changed files in this pull request and generated 4 comments.

Files not reviewed (4)

man/random_ggplot.Rd: Language not supported
man/random_mock.Rd: Language not supported
man/random_text.Rd: Language not supported
man/shinipsum-package.Rd: Language not supported

 random_ggplot <- function(type = c("random", "point", "bar",
                                   "boxplot","col", "tile",
                                   "line", "bin2d", "contour",
                                   "density", "density_2d", "dotplot",
                                   "hex", "freqpoly", "histogram",
-                                   "ribbon", "raster", "tile",
-                                   "violin")) {
+                                   "ribbon", "raster",
+                                   "violin", "ts")) {


      "ribbon" = sample(140:141, 1),
      "raster" = sample(150:151, 1),
      "tile" = sample(160:161, 1),
-      "violin" = sample(170:171, 1)
+      "violin" = sample(170:171, 1),
+      "ts" = sample(180:182, 1)


  stop_if_all(
    list(nchars, nwords),
    ~ !is.null(.x),
-    "You can't chose both nchars and nwords"
+    "You can't choose both nchars and nwords"
  )


+  the lorem-ipsum corpus on repeated calls (#3).
+* Fixed a `ggplot2` deprecation warning in `random_ggplot("line")`
+  (`size` -> `linewidth`) (#13).
 * Added a `NEWS.md` file to track changes to the package.


- new type 'ts' (alias 'timeseries') with 3 Date-axis variants (economics unemploy/psavert, AirPassengers) - fix geom_line(size=) -> linewidth deprecation warning (#13) - add coverage for random_ggplotly() - TDD: tests/testthat/test-ggplot-ts.R written first (red), then implementation

- random_mock(widths, heights, labels) returns a htmltools::tagList of bordered bootstrap columns to sketch a Shiny UI zoning - new Imports: htmltools - declare psavert/passengers globals, ignore dev/ in build - TDD: tests/testthat/test-mock.R written first (red), then implementation

- random_text(..., var = N): var = 1 keeps the historical output, higher values return distinct (wrapping) slices of the lorem corpus - TDD: tests/testthat/test-text-var.R written first (red), then implementation

Addresses pr-reviewer nits on #16: explicit heights guard with tests, and distinct wrapping default labels via random_text(var = i).

… Suggests random_ggplotly() with no arg could build a density_2d/hex variant whose rendering needs MASS/hexbin, absent on minimal CI images. Pin the test to package-free geom types, and declare MASS + hexbin in Suggests so the density_2d / hex types of random_ggplot() work (also clears the --as-cran test ERROR).

- typo 'chose' -> 'choose' in random_text() error message - rewrite random_text() var>1 word-indexing without ambiguous +/: precedence - de-duplicate 'tile' in random_ggplot() type choices (was biasing the random pool) + add a no-duplicates test

Copilot

Pull request overview

Copilot reviewed 13 out of 17 changed files in this pull request and generated 3 comments.

Files not reviewed (4)

man/random_ggplot.Rd: Language not supported
man/random_mock.Rd: Language not supported
man/random_text.Rd: Language not supported
man/shinipsum-package.Rd: Language not supported

+    } else {
+      n <- length(shinipsum::lorem_words)
+      start <- (offset + (var - 1L) * nwords) %% n
+      idx <- ((start + seq.int(0L, nwords)) %% n) + 1L


      "histogram" = sample(130:131, 1),
      "ribbon" = sample(140:141, 1),
      "raster" = sample(150:151, 1),
      "tile" = sample(160:161, 1),
-      "violin" = sample(170:171, 1)
+      "violin" = sample(170:171, 1),
+      "ts" = sample(180:182, 1)


+  the lorem-ipsum corpus on repeated calls (#3).
+* Fixed a `ggplot2` deprecation warning in `random_ggplot("line")`
+  (`size` -> `linewidth`) (#13).
 * Added a `NEWS.md` file to track changes to the package.


- merge the duplicate (unreachable) 'tile' switch branch: sample(c(40:41, 160:161), 1) so variants 160/161 are reachable - drop the stale 'Added a NEWS.md file' bullet (NEWS.md already exists)

VincentGuyader · 2026-05-12T12:17:49Z

Thanks @copilot — re-review handled in 3307c48:

"tile" duplicated in formals()$type and the "chose" → "choose" typo: these were already fixed in the previous commit (81c4f58); the re-review was looking at the earlier state.
R/Plot.R — unreachable "tile" = sample(160:161, 1) switch branch: merged into a single branch — "tile" = sample(c(40:41, 160:161), 1) — so the 160/161 variants are now reachable (and exercised by the 100-iteration random_ggplot() test, with 0 zero-coverage lines in Plot.R).
NEWS.md — "Added a NEWS.md file" bullet: removed (the file already exists; the bullet was inherited from master and not informative).

CI is fully green (6/6, test-coverage included after the rebase on the merged #14).

) The var>1 word branch produced nwords+1 indices. Use seq_len(nwords) so exactly nwords words come back; var==1 stays byte-for-byte the legacy output. Add a word-count test for var>1.

VincentGuyader · 2026-05-12T12:31:15Z

Re-review @copilot — 717dfce:

NEWS.md "Added a NEWS.md file" bullet and the duplicate "tile" switch branch: both already fixed in 3307c48 — the re-review ran against an earlier commit. The current NEWS.md has no such bullet, and R/Plot.R has a single "tile" = sample(c(40:41, 160:161), 1) branch (so 40-41 and 160-161 are all reachable; Plot.R has 0 zero-coverage lines).
R/Text.R — var > 1 returned nwords + 1 words: fixed — idx <- ((start + seq_len(nwords) - 1L) %% n) + 1L, so it now returns exactly nwords words for any var > 1, with a word-count regression test. (var == 1 is intentionally left byte-for-byte identical to the legacy output — the legacy 1 + offset:nwords + offset slicing is its own pre-existing off-by-one, out of scope for this PR.)

CI: green.

VincentGuyader added a commit that referenced this pull request May 12, 2026

refactor(random_mock): validate heights, use var-based auto labels

ffd3b2e

Addresses pr-reviewer nits on #16: explicit heights guard with tests, and distinct wrapping default labels via random_text(var = i).

VincentGuyader requested a review from Copilot May 12, 2026 08:32

Copilot started reviewing on behalf of VincentGuyader May 12, 2026 08:33 View session

Copilot AI reviewed May 12, 2026

View reviewed changes

VincentGuyader mentioned this pull request May 12, 2026

feat: random_image_ext() — Lorem Picsum images for Shiny prototyping (closes #8) #17

Open

VincentGuyader requested a review from Copilot May 12, 2026 10:57

Copilot started reviewing on behalf of VincentGuyader May 12, 2026 10:58 View session

Copilot started work on behalf of VincentGuyader May 12, 2026 11:02 View session

Copilot AI reviewed May 12, 2026

View reviewed changes

Copilot finished work on behalf of VincentGuyader May 12, 2026 11:05

VincentGuyader added 8 commits May 12, 2026 13:12

feat(random_text): add 'var' argument for text variants (#3)

b488748

- random_text(..., var = N): var = 1 keeps the historical output, higher values return distinct (wrapping) slices of the lorem corpus - TDD: tests/testthat/test-text-var.R written first (red), then implementation

docs: update NEWS.md

57a5163

refactor(random_mock): validate heights, use var-based auto labels

253a195

Addresses pr-reviewer nits on #16: explicit heights guard with tests, and distinct wrapping default labels via random_text(var = i).

docs(dev): update autonomous session log (CI green)

7be6ba1

fix: address Copilot review (#16)

81c4f58

- typo 'chose' -> 'choose' in random_text() error message - rewrite random_text() var>1 word-indexing without ambiguous +/: precedence - de-duplicate 'tile' in random_ggplot() type choices (was biasing the random pool) + add a no-duplicates test

VincentGuyader force-pushed the feat/text-variants-and-ts-plot branch from dce10a3 to 81c4f58 Compare May 12, 2026 11:12

VincentGuyader requested a review from Copilot May 12, 2026 12:09

Copilot started reviewing on behalf of VincentGuyader May 12, 2026 12:10 View session

Copilot AI reviewed May 12, 2026

View reviewed changes

fix: address Copilot review #2 (#16)

3307c48

- merge the duplicate (unreachable) 'tile' switch branch: sample(c(40:41, 160:161), 1) so variants 160/161 are reachable - drop the stale 'Added a NEWS.md file' bullet (NEWS.md already exists)

Copilot started work on behalf of VincentGuyader May 12, 2026 12:18 View session

Copilot finished work on behalf of VincentGuyader May 12, 2026 12:20

fix: random_text(var>1) returns exactly nwords words (Copilot review #16

717dfce

) The var>1 word branch produced nwords+1 indices. Use seq_len(nwords) so exactly nwords words come back; var==1 stays byte-for-byte the legacy output. Add a word-count test for var>1.

Copilot started work on behalf of VincentGuyader May 12, 2026 12:31 View session

Copilot finished work on behalf of VincentGuyader May 12, 2026 12:33

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat: random_ggplot('ts'), random_mock(), random_text(var=) — issues #4 #1 #3#16

feat: random_ggplot('ts'), random_mock(), random_text(var=) — issues #4 #1 #3#16
VincentGuyader wants to merge 10 commits into
masterfrom
feat/text-variants-and-ts-plot

VincentGuyader commented May 12, 2026

Uh oh!

VincentGuyader commented May 12, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

VincentGuyader commented May 12, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI left a comment

Uh oh!

VincentGuyader commented May 12, 2026

Uh oh!

VincentGuyader commented May 12, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

		@@ -21,18 +26,38 @@ random_text <- function(nchars = NULL, nwords = NULL, offset=0){
		~ !is.null(.x),
		"You can't chose both nchars and nwords"

Conversation

VincentGuyader commented May 12, 2026

Summary

Test plan (TDD, red first)

Uh oh!

VincentGuyader commented May 12, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

VincentGuyader commented May 12, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

VincentGuyader commented May 12, 2026

Uh oh!

VincentGuyader commented May 12, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants